4128
Vol. 52
E. P. KOHLER AND E. M. NYGAARD
The tube was placed in a water-bath a t 100" and observations made every two or three
minutes. With every arsonic acid examined a characteristic precipitate of the corresponding arseno compound (with perhaps some arsine oxide) was obtained within
twenty minutes. When a solution of arsenic trioxide in dilute hydrochloric acid was
treated with hypophosphorous acid under these conditions, no precipitate was obtained.
Table I summarizes the results obtained in testing a series of arsonic acids. I n two cases
it was found more satisfactory to dissolve the arsonic acid in hypophosphorous acid and
omit the water.
TABLE
I
ACTIONOF HYPOPHOSPHOROUS
ACID
Arsonic acid
Solvent
1 p-Hydroxyphenylarsonic acid
2 Phenylarsonic acid
3 Mono-sodium salt of p-arsonophenoxyethanol
4 p-Arsanilic acid
5 Tryparsamide
6 3-Nitro-4-aminophenylarsonic
acid
H20
HzO
H?O
(a) HzO
(b) HIPOX
H20
Minutes for
reduction
4
1
1
19
13
2
2
Color of
precipitate
Red
White
Pale yellow
Red
Solid red mass
Red
Golden-yellow
Summary
1. A rapid qualitative test of arsonation for aromatic compounds has
been suggested.
2. A new type of arsonic acid, the arsonated aromatic aldehyde, has
been developed.
3. The preparation of three arsonated aromatic aldehydes, 3-nitro-4arsonobenzaldehyde, m-arsonobenzaldehyde and p-arsonobenzaldehyde has
been described. p-Arsonobenzaldehyde was prepared by a new and
independent synthetic method.
EVANSTON,
ILLINOIS
[CONTRIBUTION FROM THE CHEMICAL
LABORATORY
O F HARVARD
UNIVERSITY ]
THE REACTION BETWEEN HIGHLY PHENYLATED COMPOUNDS
AND ORGANIC MAGNESIUM COMPOUNDS
BY E. P. KOHLER
AND E. M. NYGAARD
RECEIVED
AUGUST7, 1930
PUBLISHED
OCTOBER 6 , 1930
A study of the reaction between a,p-unsaturated ketones and organic
magnesium compounds, made many years ago, included every member
except the last of a series beginning with acrolein and ending with its most
highly phenylated substitution product, tetraphenyl propenone. In view
of the increasing interest in highly phenylated compounds and especially of
our own results with highly phenylated unsaturated nitro compounds, it
was desirable to complete the study of this series.
We have, therefore, investigated the behavior of the completely phenylated unsaturated ketone I both toward methyl magnesium iodide and
Oct., 1930
REACTIONS OF HIGHLY P H E N Y L A T E D COMPOUNDS
4129
toward phenyl magnesium bromide. And since the results were rather
surprising, we also included in the investigation the behavior of the corre11.
sponding saturated ketone, cu,P,p-triphenylpropiophenone,
(CeHs)zC=C (CeH6)COC6Hs
I
(CcH5)*CHCH(CsH5)COCeHs
I1
We found that the saturated ketone reacts much more readily than we
had anticipated, both with methyl magnesium iodide and with phenyl
magnesium bromide. Indeed the reaction with methyl magnesium iodide
proceeds rapidly in ethereal solution a t the ordinary temperature, and the
yield of the tertiary alcohol to be expected is practically quantitative.
The structure of the product was established by treating it with a dehydrating agent and ozonizing the resultant hydrocarbon. The oxidation
product was the original ketone
(CeH,),CHCH (CeH6)COCeHs +(CeH5)zCHCH ( CeHs)C (CH3)(CRHO)
OH +
I1
I11
(C~HE.)ZCHCH
(e&) c (CeH5)=CHz +(CsHs)zCHCH(CsH5)COC&
IV
With phenyl magnesium bromide the saturated ketone reacts less readily,
but this reaction can likewise be completed in ether, a t the ordinary temperature. In this case also the principal product is the tertiary alcohol but
the yield is not nearly so good, doubtless because of the ease with which the
highly phenylated carbinols undergo cleavage to a mixture of simpler
products. The proof of the structure of the Grignard product was obtained
by oxidation. The carbinol was readily oxidized by chromic acid in glacial
acetic acid, and yielded two moles of benzophenone to one of benzoic acid.
11
+ (CeHs)*CHCH(CeH6)C(CBHS),oR+ 2(CsHa)zCO +
C6HaCOsH
V
In sharp contrast with the saturated ketone, the corresponding unsaturated compound does not react a t all with either of the Grignard reagents a t the ordinary temperature, and it was recovered even after prolonged boiling. In order to secure any reaction it was necessary to replace
most of the ether with benzene and operate a t a higher temperature.
The reaction between the unsaturated ketone and methyl magnesium
iodide gave two solid products and a small quantity of oil. The composition of the principal product indicates that it is formed as a result of the
addition of one molecule of methyl magnesium iodide to the unsaturated
ketone. The second product, always formed only in relatively small
quantities, is a hydrocarbon which is doubtless formed from the primary
product by loss of water.
A quantitative examination of the behavior of the primary product
toward methyl magnesium iodide showed that it reacts with only one mole
of the reagent and liberates one mole of gas; it therefore probably is a
4130
VOl. 52
E. P. KOHLER AND E. M. NYGAARD
hydroxyl compound. It reacts with ozone and is in part converted into a
peroxide, and in part oxidized to benzophenone and other products.
These transformations indicate that the substance is a tertiary alcohol
formed by addition to the carbonyl group. The results are, however, not
quite conclusive because it was impossible to prove the presence of methyl
benzoin in the ozonization products.
(CsHdiCO
+ C6"sCOC(CsHs) (CHdOH
More definite proof of the structure of the substance was obtained by
studying the hydrocarbon that is formed in the same reaction. This
hydrocarbon can be obtained a t will by heating the primary product with
acetic anhydride and sodium acetate. When it is gently oxidized with
chromic acid, it first combines with two atoms of oxygen and forms a diketone and when this diketone is oxidized more intensively it passes into
ortho dibenzoyl benzene. The hydrocarbon is, therefore, methyl triphenyl
indene and the primary product must be the carbinol, VI.
(CsHs)&=C(CsHs)C(CHs)(CsHs)OH
----f
@z:
CCeHdCH,)
+
VI1
VI
n C ( c s H s ) (CH,)COCsHs
0CC)Cs"
d
VCOCsHr
VI11
It proved to be far more difficult to establish the course of the reaction
with phenyl magnesium bromide because, although the conditions under
which the experiments were conducted were varied in every conceivable
manner, the almost invariable result was a mixture of heavy oils from
which but little solid product could be extracted. It was not until we
adopted the device of decomposing the magnesium compound with acid at
the lowest possible temperature and subjecting the product to the action
of air or oxygen that we began to get solids in sufficient quantity for
examination.
It is known from earlier experiments,' that when the enolic modifications
which are obtained by 1,Caddition to such unsaturated ketones as contain
a hydrocarbon residue in the a-position are exposed to oxygen they form
peroxides. This procedure in our case likewise resulted in the addition of a
molecule of oxygen but surprisingly enough the product had a deep yellow
color. We were therefore confronted with the fact that we had started
with a colorless unsaturated compound, added first the equivalent of a
Kohler, Am. Chem. J.,36, 181 (1906).
Oct., 1930
REACTIONS OF HIGHLY PHENYLATED COMPOUNDS
4131
molecule of benzene and then a molecule of oxygen, and had ended with a
deep yellow product which to all appearances was more highly unsaturated
than our original substance. The conclusion appeared inevitable that a
benzene ring had become involved in the reaction. The transformations
of the product confirmed this conclusion.
When the yellow peroxide was treated with cold sodium methylate,
which usually cleaves these enolic peroxides, it lost oxygen and water and
passed into a new yellow compound which had the composition CaaHuO.
This new yellow compound was subjected to the action of ozone, which, as
usual, added two atoms of oxygen and gave two products-benzophenone
and a third yellow substance with the composition CZOH~~OZ.
Inasmuch as a yellow compound that is formed by ozonization is most
likely to be an a-diketone, we treated this yellow ozonization product with
ortho phenylene diamine and found that it readily formed a glyoxaline
derivative. Ortho diketones are most cleanly oxidized with alkaline hydrogen peroxide, which cleaves them between the two carbonyl groups. With
this reagent our third yellow product gave, quantitatively, benzoic and
ortho phenyl benzoic acids.
The organic degradation products of our peroxide are, therefore, benzophenone, benzoic acid and ortho phenyl benzoic acid. The ortho phenyl
benzoic acid proves that one of the phenyl groups participates in the reaction, and when all the degradation products are reassembled in their
proper order they show quite conclusively which of the phenyl groups is
involved, the long series of transformations being represented as follows
IX
0-0
I 1 CH-CH
(CeH&C=C-C-C<
\CH
CsHsOH
1 I
CH(C~H~)---CH/
X
(CeHs)zCO
CeHaCOCOCeH&Hs ( 0 )
XI1
+
+(GHs)2C=C-COc>
'
CsHS
XI
+
cas
+ CeHpCOzH 4- Ce.HsCsH&OzH
(0)
In addition to the 1,4-addition product IX, there is formed a relatively
small quantity of an unstable colorless compound. This substance liberates gas from methyl magnesium iodide and is capable of forming an acetate. Since it is colorless it cannot be the enol IX, but it might be either a
rearrangement product of the enol (XIV) or the 1,2-addition product (XV)
(CeH6)zC=C(CoHs) CHOHCeHdCsHp ( 0 )
XIV
(C~H~)ZC=C(C~HK)C(C&I~)~OH
XV
In order t o distinguish between these two possible formulas we oxidized
the acetate and found that one mole of the substance gives one mole of
4132
E. P. K O H L E R AND E. M. NYGAARD
Vol. 52
benzoic acid and two moles of benzophenone, proving that the substance
must be the 1,2-addition product. Our total yield of definite products is
unsatisfactory; a t best, we account for only 70% of the unsaturated ketone.
But whether this low yield is due to the formation of other substances or the
difficulty of isolating these unstable compounds we have been unable to
ascertain.
Taken in their entirety, our results show that the most highly phenylated
a,@-unsaturatedketone, like most other members of the series, is capable of
forming both 1,2- and 1,4-addition products with organic magnesium
compounds. Here, however, the resemblance ceases. In all other cases in
which both types of addition occur, the relative quantity of the 1,4-addition
product is always larger with alkyl than with aryl magnesium compounds.
In our case these relations are completely reversed. Methyl magnesium
bromide gives no 1,4-addition product, phenyl magnesium bromide but
little of the 1,2-product.
A much more interesting difference, however, is represented by the fact
that it is not, as in the case of all other members of this series, the ethylenic
system but a phenyl group that is involved in l,.l-addition. The firstand only other-case of an addition of this type was reported last year.
The two cases are not dissimilar. Gilman and his co-workers2found that
benzophenone ani1 reacts with phenyl magnesium bromide only a t a relatively high temperature, and that the reaction then consists in L4-addition
and involves one of the phenyl groups
CeHaN=C(Ce")z
--f
C~H~N-C=C~HIC~H
( 0~
)
I
I
MgBrCeH6
These two cases serve to diminish to some extent the difference between
aromatic and ethylenic systems. It is clear that a phenyl group will
participate in a reaction involving 1,4-addition when the hindrance to
every other reaction is sufficiently great. It is far from clear, however,
why there should be so much more hindrance to 1,2-addition in the unsaturated than in the saturated ketone.
Experimental Part
I. Preparation of Materials
No better methods were found for preparing either the saturated or the
unsaturated ketones, but the yields have been improved greatly by modifying the procedure a t nearly every stage. The steps involved in the preparation of the saturated ketone are shown in the following series of formulas
C&CHOHCOCsHs
+CsHaCHeCOCaHs +CaHaCHC1CH(CaHs)COCeHs ----f
CeHsCH=C (CeHs)COCgHs ----f (CeHs)zCHCH(CeH6)COCeHs
2
Gilman. Kirby and Kinney, THISJOURNAL, 51, 2252 (1929).
Oct., 1930
R E A C T I O N S O F H I G H L Y P H E N Y L A T E D COMPOUNDS
4133
For the reduction of benzoin to desoxybenzoin we used zinc and glacial
acetic acid as recommended by S u d b ~ r o u g hbut
, ~ by introducing an effective mechanical stirrer and by raising the temperature from that of the
steam-bath to that of the boiling point of the liquid we not only greatly
decreased the time of heating but improved the yield as well. The success
of this method depends almost entirely on the quality of the zinc dust.
With the finely powdered zinc which is frequently sold as zinc dust the yield
is small. Instead of separating a part of the product by crystallization
we found it advantageous to subject the entire product to distillation
under diminished pressure.
Desoxybenzoin.-Two lots, each of 346 g. of benzoin, 600 g. of glacial acetic acid
and 105 g. of zinc dust contained in a flask provided with a mechanical stirrer and an air
condenser, were heated in an oil-bath a t the boiling point of the liquid until a sample no
longer gave a flocculent precipitate when i t was dropped into water (four t o five hours).
The benzoin was first dissolved in the acid heated t o about 115", then the zinc dust
was added as rapidly as possible without causing frothing, and finally the temperature
was raised t o the boiling point.
The solution was poured into hot water, the mixture boiled until the oil collected
on the bottom, then cooled and filtered. The solid was washed with water and dissolved
in alcohol. After removal of the zinc, the alcoholic solution was distilled under ordinary
pressure until free from alcohol, then fractionated under diminished pressure. The
yield was 534 g. of desoxybenzoin, equal t o 82'%.
p-Chloro Benzyl Desoxybenzoin.-The desoxybenzoin obtained in the preceding
experiment was dissolved in 350 g. of benzaldehyde, the solution cooled in a freezing
mixture, and saturated with hydrogen chloride which was passed into it for seven t o
eight hours. After standing overnight the resulting solid cake was broken up under
alcohol, washed-first with alcohol, then with ether-and dried. The yield was 622 g.
of chloro compound from 534 g. of desoxybenzoin, equal t o 71%.
Benzal Desoxybenzoh-A suspension of 250 g. of p-chloro benzyl desoxybenzoin,
200 g. of fused potassium acetate and 65 g. of dry sodium carbonate in one liter of
methyl alcohol was boiled with constant vigorous mechanical stirring for three hours,
then cooled and filtered. The solid was washed with sufficient water to remove inorganic salts completely. I t s melting point was 100 ', only 1' below that of the pure
substance, which was obtained by one crystallization from ether. The yield is 180 g.equivalent t o 81%.
The Saturated Ketone (II).-To a solution of phenyl magnesium bromide prepared
from 10 g. of magnesium, 50 g. of finely ground benzal desoxybenzoin was added as fast as
it dissolved. The mixture was heated for a short time, then cooled and decomposed with
ammonium chloride and ammonia in the usual manner. It yielded 61 g. of pure ketone
equivalent t o 95.8%.
The saturated ketone cannot be brominated. The route to the unsaturated ketone,
therefore, follows the course
CsH5CH=C(C6H6)COC6H5---f (CsH5)2CHC(CsHs)=C(OMgBr)C~Hs
+
( C ~ H S ) ~ C H C ( C ~ H ~ ) B ~+
C O C( C
~H
~H
~~)ZC=C(CBH~)COC~H~
For the purpose of preparing the a-bromo compound, 100 g. of finely ground benzal
desoxybenzoin was added t o a solution of phenyl magnesium bromide made from 20 g.
of magnesium. The mixture was heated for a short time, then cooled in a freezing mixa
Sudborough, J . Chem. SOC.,71, 219 (1897).
4134
E. P. KOHLER AND E . M. NYGAARD
VOl. 52
ture. To it was added, with constant stirring and very slowly, t o avoid local rise in
temperature, 120 g. of bromine which had been dried over phosphorus pentoxide. The
mixture was poured into ice and hydrochloric acid and shaken until solid began t o separate. The ethereal layer was then diluted with petroleum ether. The resulting solid
was washed with water, a mixture of ether and petroleum ether and finally petroleum
ether alone. The yield of a-bromo compound was 133 g., equivalent t o 87%.
The Unsaturated Ketone.-A suspension of 130 g. of the a-bromo compound and an
equal weight of potassium acetate in 650 g. of dry ethyl alcohol was boiled for two hours
with sufficiently rapid stirring t o prevent bumping. Most of the unsaturated ketone
separates on cooling, mixed with potassium chloride. After washing and drying it is
M c i e n t l y pure for most purposes. The remainder of the product is obtained by distilling off most of the alcohol, diluting the residue with water and extracting with ether.
This is less pure and needs t o be recrystallized from chloroform and alcohol. The yield
was 94.5 g., equal t o 89%.
11. Experiments with the Saturated Ketone
2,3,4,4,-Tetraphenyl Butanol-2 (III).-Fifteen grams of the finely ground saturated
ketone was added t o a solution of phenyl magnesium bromide made from 4 g. of magnesium. There was no perceptible warming and no change in color but the solid gradually dissolved. After remaining for two hours a t the ordinary temperature, the clear
solution was decomposed with iced ammonium chloride in the usual manner. The
washed and dried ethereal layer, on evaporation, left a solid. This was recrystallized
from methyl alcohol. The yield was 12.5 g. of pure product and 2 g. of material, largely
composed of the same substance as the principal product.
Anal. Calcd. for GBHI~O:C, 88.8; H, 6.9. Found: 88.4; H, 7.0.
The carbinol is readily soluble in all common organic solvents except petroleum
ether. It crystallizes in long, flat needles and melts at 140". It appears t o be stable
in the air but its melting point gradually drops when it is kept over sulfuric acid in a
vacuum desiccator.
2,3,4,4-Tetraphenyl Butene (IV).-A solution of one gram of the hydroxyl compound in the requisite quantity of glacial acetic acid containing two drops of sulfuric
acid was set aside for twelve hours. It deposited 0.71 g. of a colorless solid and when the
acid was pumped off it left 0.2 g. more of the same substance. The substance is quite
readily soluble in all common organic solvents including petroleum ether. It was recrystallized from methyl alcohol, from which it separated in needles melting at 104-106
Anal. Calcd. for CI~HM:C, 93.3; H, 6.7. Found: C, 93.2; H, 6.9.
Ozonization.-A current of ozonized oxygen containing about 6% of ozone was
passed through a solution of 2 g. of the hydrocarbon in ethyl bromide for two hours. A
considerable quantity of a crystalline solid separated from the solution during the passage
of the ozone and more of the same substance was obtained when the ozonized solution
was manipulated in the usual manner. The solid had the melting point of the saturated
ketone-tetraphenyl propanone-and it caused no depression of the melting point
when it was mixed with this ketone. The saturated ketone was likewise obtained
when the hydrocarbon was oxidized with permanganate.
1,1,2,3,3-Pentaphenyl Propanol-1 (V).-A solution of 5 g. of the saturated ketone in
dry benzene was added t o a solution of phenyl magnesium bromide which had been
made from 1.4 g. of magnesium and cooled in a freezing mixture. After the addition,
the solution was stirred at the ordinary temperature for a n hour. A test made a t this
time still showed much unchanged ketone. The mixture was therefore set aside for
seven hours, then decomposed with iced acid in the usual manner. I n addition to much
oil, it gave 3.0g. of a solid melting a t 158-159 '-a yield of about 33%.
'.
Oct., 1930
RGACTIONS OF HIGHLY PHENYLATED COMPOUNDS
4135
All attempts to reduce the quantity of oil and increase the yield of the solid carhinol
were unsuccessful. Lowering the temperature and stopping the reaction sooner resulted
in mivtures of products and unchanged substance which were difficult to handle, and
decomposing with ice and ammonium chloride in place of hydrochloric acid had no effect
on the yield.
The solid product was purified by crystallizing it from methyl alcohol, from which
it separates in the form of small white prisms melting a t 159'.
Anal. Calcd. for Ca3Hx8O:C, 90.0; H.6.4. Found: C, 89.5; H, 6.8.
A test by the Zeisel method showed that the substance has no alkoxy1 group, and a
quantitative examination of its reaction with a solution of methyl magnesium iodide in
iso-amyl ether showed that it reacts with one mole of the reagent and liberates one mole
of gas.
Oxidation.-All attempts to eliminate water from the hydroxyl compound led to
unmanageable oils. The compound was therefore oxidized with excess of chromic acid
in hot glacial acetic acid. To this end one gram of the substance in glacial acetic acid
was treated with excess of chromic acid and heated on a steam-bath for two hours.
It yielded 0.73 g bf benzophenone and 0.16 g. of benzoic acid, instead of 0.83 g. of the
former, 0.2 g. of the latter assuming that one mole of substance gives two moles of the
ketone and one mole of the acid.
111. Experiments with the Unsaturated Ketone
A. Reaction with Methyl Magnesium Iodide
The reaction between methyl magnesium iodide and the unsaturated ketone was
carried out a great many times. In the early experiments we became convinced that
the reaction would not run in ether alone, no matter how prolonged the boiling. We
then turned to benzene and found that the minimum temperature at which the reaction
would run a t an appreciable rate was about 50'.
In subsequent experiments, therefore, we used four equivalents of the reagent,
replaced the ether with benzene, heated for periods ranging between forty minutes and
ten hours and decomposed sometimes with ice and acid, sometimes with ice and ammonium chloride. The yields of carbinol varied from 53.6 to 63.4%. I n addition to the
carbinol the product usually contained small quantities of its dehydration productmethyl triphenyl indene. No evidence was found for any 1,4-addition product. The
lack of color showed conclusively that no phenyl group was involved in the reaction,
and the inability t o obtain a peroxide was good even if inconclusive evidence that no
other type of 1,4-addition had occurred.
2,3,4,4-Tetraphenyl Butenol-2 (VI).-A solution of 10 g. of the unsaturated ketone
in 60 cc. of dry benzene was added rapidly to a solution of methyl magnesium iodide
which had been prepared in an atmosphere of dry nitrogen from 2.72 g. of magnesium.
The mixture was heated to 60" for ten hours while a current of dry nitrogen was passed
through the apparatus. After decomposing the magnesium compounds with ice and
ammonium chloride in the usual manner, the benzene-ether layer was washed, dried
and evaporated in a current of dry air. I t deposited a solid which after washing with
petroleum ether melted a t 96.5'. therefore was pure carbinol. The petroleum ether
washings yielded 0.8 g. of methyl triphenyl indene. The rest was oil.
The carbinol dissolves freely in all common organic solvents except petroleum ether.
It crystallizes well from ether and petroleum ether and it can also be recrystallized from
methyl alcohol and water. I t separated in white plates and melts a t 96.5'.
Anal. Calcd. for CzsHntO: C, 89.4; H, 6.4. Found: C, 89.5; H, 6.7.
Proof of Hydroxyl Group.-One mole of the substance reacted with one mole of
methyl magnesium iodide and liberated one mole of methane.
4136
E. P. K O H L E R A N D E. M. NYGAARD
Vol. 52
Proof of Double Linkage.-The carbinol was ozonized in ethyl bromide with
ozonized oxygen containing 6% ozone. The product, decomposed with ice in the usual
manner, and then set aside, slowly deposited a small quantity of solid. After recrystallization from alcohol the solid melted with decomposition a t about 148". Its composition corresponds to that of a peroxide of the carbinol and it flashed when it was
heated rapidly.
Anal. Calcd. for C28H240.02: C, 82.4; H, 5.9. Found: C, 82.3; H, 6.0.
The oil that was left after the removal of the peroxide was distilled with steam.
The oil in the distillate formed an oxime which melted a t 141 and a phenylhydrazone
melting a t 138O , therefore was benzophenone.
Conversion into Methyl Triphenyl Indene.-A solution of one gram of the carbinol
and the same quantity of sodium acetate in acetic anhydride was boiled for five hours,
then poured into water and made alkaline with excess of sodium carbonate. The alkaline solution was extracted with ether, the ethereal layer washed, dried and evaporated.
It left a solid which after washing with a little ether proved t o be pure methyl triphenyl
indene. The yield was 0.67 g.
Methyl Triphenyl Indene (VII).-The indene derivative, as has been shown, can be
obtained in an excellent yield by heating the carbinol with a suitable condensing agent.
It was generally obtained in small quantities as a by-product of the Grignard reaction.
In one experiment, 50 g. of the ketone gave 12 g. of the pure indene derivative. Whether
this uncommonly large yield was formed during the reaction or during the decomposition
of the magnesium derivatives with hydrochloric acid could not be ascertained because
we were unable t o duplicate the result. The indene crystallizes from absolute ethyl
alcohol in the form of prisms melting a t 118".
Anal. Calcd. for C2~H22: C, 93.9; H , 6.2; mol. wt., 358. Found: C, 93.8; H,
6.2; mol. wt., 320.
Oxidation: a,a,a-Benzoyl-methyl-phenyl-ortho-tolylPhenyl Ketone (VIII).-A
solution of 1.5 g. of the hydrocarbon and 1.2 g. of chromic acid in glacial acetic acid was
warmed until oxidation began a t 50-60". The temperature was held a t this point for
fifteen minutes. The solution was then diluted with water and extracted with ether.
The washed and dried ethereal solution was concentrated and diluted with petroleum
ether, whereupon it deposited a solid. The solid was recrystallized from anhydrous
ether, from which it separated in white prisms melting a t 182'.
Anal. Calcd. for C2~H2202:C, 86.2; H , 5.6. Found: C, 86.4; H , 5.2.
Ortho Dibenzoyl Benzene.-The oily mother liquors from the diketone were oxidized
with a larger quantity of chromic acid. The principal product was o-dibenzoyl benzene,
which was identified by comparison with a sample on hand. Ortho dibenzoyl benzene
was also one of the products formed when the oily by-products of the Grignard reaction
were oxidized with excess of chromic acid.
B. Reaction with Phenyl Magnesium Bromide
A series of preliminary experiments showed that the unsaturated ketone does not
react with phenyl magnesium bromide in boiling ether and that when the reaction was
carried out in boiling benzene no solid products could be isolated. We used, therefore, a
large excess of the reagent-usually four equivalents-and operated a t temperatures
between 45-60", The procedure most favorable for securing the peroxide was the least
favorable for isolating the unstable 1,2-addition product. Moreover, the acetate of the
1,2-addition product was most easily obtained directly from the magnesium derivative
( C ~ H ~ ) ~ C = C ( C B H ~ ) C O---f
C B H(CsH5)2C=C(CsH5)C(CsH6)20MgBr
~
+
(Ce")zC=c(Ce") C(CsH5)zOCOCH,
Oct., 1930
REACTIONS OF HIGHLY P H E N Y L A T E D COMPOUNDS
4137
The Peroxide (X).-A solution of 15 g. of the ketone in 90 cc. of benzene was added
drop by drop to an ethereal solution of phenyl magnesium bromide that was made from
4.08 g. of magnesium. Each drop of the solution produced a momentary red color,
which turned to yellow, but by the time all of the ketone had been added, the color of the
solution was dark red t o brown. Ether was boiled off until the boiling point of the solution reached 50-52 '.
After the solution had been boiled a t 50-52" for an hour, it was cooled and poured
into iced hydrochloric acid contained in a separating funnel. The mixture was shaken
vigorously, the acid layer drawn off and the ethereal layer rapidly washed with ice water.
The washed ethereal solution was then transferred to a suction flask which was cooled
with ice while a rapid current of air was drawn through it. A yellow precipitate soon
began t o separate, This was collected on a filter when the solution became too thick;
a second crop was obtained by diluting the oily residue with ether and repeating the
operation, and a third by treating the oil with ether and petroleum ether. The total
yield was 7.3 g., equivalent t o 37.370.
The peroxide was recrystallized from ether, from methyl alcohol, and from chloroform, but it is doubtful that this increased its purity. Its melting point varied with the
mode of heating, the highest being 186.5'.
Anal. Calcd. for C33H2603: C, 84.3; H, 5.5. Found: C, 84.0; H, 6.0.
In the machine one mole of the compound consumed three moles of methyl magnesium iodide and liberated one mole of gas. The peroxide crystallizes in long yellow
prisms.
Ozonization.-The peroxide was oxidized with ozone in the usual manner but the
results were not very instructive. It gave a considerable quantity of benzophenone,
showing that the ethylene linkage was intact but, as was t o be expected, the remainder
of the molecule was converted into highly colored oils from which only benzoic acid could
be isolated.
the perReaction with Sodium Methylate. The Unsaturated Ketone (=).-When
oxide is added to a concentrated solution of sodium methylate in methyl alcohol it
dissolves in part and forms a brilliant red solution, but the ketone begins t o separate
before solution is complete. The reaction appears to be complete in an hour, but the
mixture was usually set aside overnight. Thus 2.5 g. of the finely ground peroxide
was added to 20 cc. of 10% sodium methylate. The mixture was set aside for fifteen
hours, then diluted with water and extracted with ether. The washed and dried yellow
ethereal solution, on evaporation, deposited 2.45 g. of crude product. This was washed
with methyl alcohol and further purified by recrystallization from the same medium.
The yield of pure ketone was 1.6 g., equivalent t o 69%.
Anal. Calcd. for Cs3H240: C, 90.8; H, 5.5; mol. wt., 436. Found: C, 90.9;
H, 5.6; mol. wt., 412.
The color of the ketone is a lighter yellow than that of the peroxide. It crystallizes
in long plates and melts a t 138".
Ozonization: Ortho Phenyl Benzil (XII) -A current of ozonized oxygen containing
about 6% of ozone was passed through a solution of one gram of the ketone in ethyl bromide for one and one-half hours. The ozonide was decomposed in the usual manner
and the solvent removed. The residue was an oil which partially solidified when it was
rubbed with alcohol. The revulting solid was recrystallized from alcohol, from which it
separated in pale yellow plates melting a t 80 '.
Anal. Calcd. for CzuHlaO~:C, 83.9; H, 4.9. Found: C, 83.5; H, 5.3.
The filtrate from the diketone was distilled with steam and the distillate extracted
with ether. The ethereal solution contained benzophenone, which was identified by
converting it into phenylhydrazone and comparing this with a sample on hand.
4138
E. P. KOHLER AND E. M. NYCAARD
Vol. 52
N
The Glyoxaline Derivative,
axc6"
.-An
alcoholic solution of
CCsH,GHdo)
0.1 g. of the diketone, 0.09 g. of ortho phenylene diamine hydrochloride and an equivalent quantity of sodium acetate was boiled for a short time, then diluted with water
and extracted with ether. The ethereal extract was freed from excess of the diamine
by shaking with dilute acid, then washed, dried and evaporated. It deposited a colorless solid. The solid was recrystallized from alcohol, which deposited it in large square
plates which melted a t 163".
Anal. Calcd. for CzsHtaNz: c , 87.1; H , 5.0. Found: C, 86.6; H , 5.1.
Oxidation of the Diketone: Ortho Phenyl Benzoic Acid.-A solution of 0.2 g. of the
diketone in methyl alcohol was treated first with excess of 15% hydrogen peroside,
then with excess of 10% sodium hydroxide. I t soon became colorless. I t was then
heated to remove the methyl alcohol, and acidified. From the acid solution ether extracted 0.23 g. of a mixture of acids which could not be separated by steam distillation,
but which was read.ly separated by sublimation. The sublimate was benzoic acid.
I t was identified as ortho phenyl benzoic acid by
The residue (0.15 9.) melted a t 113
comparison with a sample of this acid which was obtained by fusing fluorenone with
potassium hydroxide.
Triphenylvinyl Diphenyl Carbinol (XV).-A solution of 10 g. of the ketone was
added in the usual manner to a solution of phenyl magnesium bromide made from 2 7
g. of magnesium. The mixture was heated to 60" for two hours, then decomposed with
iced hydrochloric acid. The ethereal layer was washed, dried and then evaporated in a
draught. It first deposited the yellow peroxide, and later a colorless solid which was
purified by recrystallization from ether-petroleum ether; yield, 12%.
O 90.4;
:
H , 5.9. Found: C, 90.1; H, 6.0.
Anal. Calcd. for C ~ J H Z ~C,
'.
The carbinol is sparingly soluble in alcohol, moderately soluble in ether. It crystallizes in thin white plates. I t melts with decomposition, and the melting point varies
with the rate of heating. The highest melting point observed was 169'. Even the
purest samples slowly change a t the ordinary temperature, turning ultimately into
brown viscous oils, but it can be kept apparently without change a t 0'.
The Acetate (XVI).-The acetate was first obtained by boiling a solution of the
carbinol for seventeen hours with acetyl chloride and then distilling off the excess of
the chloride. It is made much more easily directly from the magnesium derivative of
the carbinol. Thus a solution of 25 g. of the ketone in benzene was added to an ethereal
solution of phenyl magnesium bromide which contained 6.8 g. of magnesium. The
mixture was kept for one hour a t 50°, then cooled in a freezing mixture. Into the cooled
solution was dropped a mixture of 33 g. of acetyl chloride and twice its volume of absolute ether. The brilliant red solution was allowed to reach the temperature of the
room, then decomposed with ice and acid in the usual manner.
The ethereal layer was thoroughly washed with water, dried and concentrated.
It yielded 10.6 g. of the same acetate that had been obtained from the carbinol. It was
pursed by recrystallization from chloroform and ether.
Anal. Calcd. for C ~ ~ H L ~C,
O 87.5;
Z : H, 5.8. Found: C, 87.5; H , 6.2.
The acetate is readily soluble in benzene and in chloroform, very sparingly soluble
in ether and in alcohol. I t crystallizes in colorless plates and melts a t 218'.
Oxidation.-A solution of one gram of the acetate and an equal weight of chromic
acid in glacial acetic acid was heated on a steam-bath for about an hour, then cooled and
Oct., 1930
THE SYNTHESIS OF THIAZOLE AMINES.
V
4139
diluted with ice and ether. The ethereal layer was shaken with sodium carbonate,
which extracted benzoic acid, then dried and evaporated. The residue when distilled
with steam yielded 0.61 g. of pure benzophenone.
Summary
When tetraphenyl propenone reacts with phenyl magnesium bromide,
the principal product is a diphenyl derivative which is formed by 1,4addition to the system -CO-ChHs.
CONVERSE
MEMORIAL
LABORATORY
CAMBRIDGE,
MASSACHUSETTS
[CONTRIBUTION FROM THE
DEPARTMENT
OF
CHEMISTRY, YALE UNIWRSITY]
THE SYNTHESIS OF THIAZOLE AMINES POSSESSING
PHARMACOLOGICAL INTEREST. V
BY W. S. HINEGARDNER~
AND T.B. JOHNSON
RECEIVED
AUGUSTa, 1930
PUBLISHEDOCTOBBR6, 1830
In Paper IV of this series, Hinegardner and Johnson2 have described the
synthesis of 2-phenylthiazole-4-ethylamine,expressed structurally by
formula I. This is a representative of a new type of aliphatic amines in
which the thiazole nucleus has been substituted for a methylene radical in
y-phenylpropylamine. It is a bridged thiazole compound of pharmacological interest and is only one of a series of compounds of its type which
may be prepared by our method of synthesis. In this paper we describe a
series of intermediate compounds which have been prepared in the developCH-S
It
HzNCHzCHzC
I
CCsHs
CsHsCH&H2NHa
HOGHICH~CH~NHI
\”
I
I1
I11
ment of a practical synthesis of 2-p-hydroxyphenylthiazole-4-ethylamine,
XIV. This latter amine bears the same relationship to tyramine I11 as the
thiazole amine I does to phenylethylamine 11. It is a very potent substance biologically and its pharmacological activity is being investigated.
The starting points for our research were sym.-dichloro-acetone and the
thioamide of anisic acid. These interact smoothly when warmed together
in alcoholic solution, giving an excellent yield of the primary halide IV.
Utilizing then the same technique as was described in our previous paper’
for the preparation of the amine I, the various transformations recorded in
Table I have been carried through successfully, leading up to the desired
amine, XIV. The experimental data establishing the constitution and
chemical identity of these various thiazoles are recorded in Table 11.
Metz Research Fellow in Organic Chemistry, 1928-1929.
* Hinegardner and Johnson,
THIS
JOURNAL, 52,,3724(1930).